In [ ]:
##2) we first observeing the data through histogram.
## 2.1 ISSUE_DATE, by day, the purpose is to see if there are some trends in the issue date, may be the issue of constructions are decreasing, or maybe there are some other patterns.
df_issuedate = copy.deepcopy(df['ISSUE_DATE'])
dt_issue, dt_issue_month,dt_issue_monthsliced, count = [], [], {}, 0
def yearcheck(dt_issue, count = 0, name = 'issue'): #check year see if in 2022 or 2023
for i in dt_issue:
if i[0:4] != '2023' and i[0:4] != '2022':
count += 1
print('Num of year out of 2022 and 2023:',count,f'\t(0 means the {name} time period is not out of year 2023)') # 0 means the time period is not out of year 2023
def datetreat(df_issuedate, dt_issue=[], dt_issue_month=[], dt_issue_monthsliced={}): #put date processes in a function for easy treat with other date data.
for i in df_issuedate: # accurate to date
dt_issue.append(i[0:10])
for i in range(len(df_issuedate)): #accurate to month
dt_issue_month.append(df_issuedate[i][2:7])
dt_issue.sort()
dt_issue_month.sort()
# print(dt_issue[0:5],'\n',dt_issue_month[0:5]) #checkline
for i in ['01','02','03','04','05','06','07','08','09','10','11','12']: #initialize the dt_issue_monthsliced
dt_issue_monthsliced['-'.join(['22',i])] = []
if i == '01' or i == '02' or i == '03' or i =='04':
dt_issue_monthsliced['-'.join(['23',i])] = []
# print(dt_issue_monthsliced) #chekline
# datetreat(df_issuedate, dt_issue=[], dt_issue_month=[], dt_issue_monthsliced) #checkline
for i in range(len(dt_issue)): #now get the 2022 and 2023 month slice
dt_issue_monthsliced[dt_issue_month[i]].append(dt_issue[i][8:10])
# print(dt_issue,'\n', dt_issue_month,'\n',dt_issue_monthsliced)
return dt_issue, dt_issue_month, dt_issue_monthsliced
dt_issue, dt_issue_month, dt_issue_monthsliced = datetreat(df_issuedate, dt_issue, dt_issue_month, dt_issue_monthsliced)
yearcheck(dt_issue, count, name = 'issue')
def month_show(dt_issue_month, name='issue'): #easy function for month_show histgram
sns.histplot(data = dt_issue_month)
plt.title("Month hist for {0}".format(name))
plt.xticks(rotation=-45)
plt.show()# we found that there's only three month's data, and the number of construction-site may be decreasing?(the date data is apprently not enough)
month_show(dt_issue_month, name = 'issue')
print("It seems that the month distribution is equal.")
for key in dt_issue_monthsliced:
sns.histplot(data = dt_issue_monthsliced[key])
plt.title(f'month {key} issue num')
plt.xticks(rotation=-45)
plt.show()
# print(key)
# print(dt_issue_monthsliced['22-01'])
print('Comparing the pictures we finally find that all days with few issue numbers seems all weekends, so that there may be fewer application than weekdays.')
Num of year out of 2022 and 2023: 0 (0 means the issue time period is not out of year 2023)
It seems that the month distribution is equal.
Comparing the pictures we finally find that all days with few issue numbers seems all weekends, so that there may be fewer application than weekdays.
